Isomerization of pentacoordinated molecules - Journal of the

Doyle Britton, and J. D. Dunitz. J. Am. Chem. Soc. , 1975, 97 (13), pp 3836–3837. DOI: 10.1021/ja00846a056. Publication Date: June 1975. ACS Legacy ...
2 downloads 0 Views 272KB Size
3836 a r e the five adamantyl systems, though isomers 7, 8, and 9 d o emerge in the correct order of stability. (iv) The strain energy of diamantane is larger than expected. Using our AHfo (8) values and the Schleyer and Allinger strain-free increments, we obtain strain energies of 15.5 and 7.7 kcal mol-’ for diamantane and adamantane, respectively. Thus the former is almost exactly twice as strained as the latter. It appears, therefore, that both force fields require some reparameterization.

Acknowledgments. W e thank I.C.I. Ltd for a fellowship (T.C.) and The Northern Ireland Department of Education for a postgraduate studentship (T.Mc0.K.). References and Notes (1) E. M. Engler, J. D. Andose. and P. v. R. Schleyer. J. Am. Chem. Soc.. 95, 8005 (1973), and references contained therein. (2) N. L. Allinger, M. T. Tribble. M. A. Miller, and D.W. Wertz, J. Am. Chem. Soc., 93, 1637 (1971). (3) R. H. Boyd, S. N. Sanwal, S. Shary-Tehrany, and D. McNally, J. Phys. Chem., 75, 1264 (1971). (4) (a) Commercial sample: (b) prepared from the appropriate bromide and methylmagnesium iodide using the Grignard coupling procedure of E. Osawa, 2. Majerski, and P. v. R. Schleyer, J. Org. Chem., 36, 205 (1971); (c) P. v. R. Schleyer and R. D. Nicholas, J. Am. Chem. SOC., 83, 186 (1961); (d) B. R. Vogt. Tetrahedron Lett., 13, 1575 (1968); (e) T. M. Gund, E. Osawa, V. 2. Williams, Jr., and P. v. R. Schleyer, J. Org. Chem., 39, 2939 (1974); T. Courtney, D. E. Johnston, M. A. McKervey. and J. J. Rooney. J. Chem. Soc., Perkin Trans 1, 2691 (1972); (f) T. M. Gund, M. Nomura, and P. v. R. Schleyer, J. Org. Chem., 39, 2987 (1974). (5) W. A. Keith and H. Mackle, Trans. Faraday Soc., 54, 353 (1958); the precision of the bomb has been considerably improved by modifications the details of which will be described elsewhere. (6) T. Clark, T. McO. Knox, H. Mackle. M. A. McKervey, and J. J. Rooney, J. Chem. SOC.,Faraday Trans 1. in press. (7) M. Mansson. N. Rapport, and E. F. Westrum, Jr., J. Am. Chem. SOC., 92, 7296 (1970). (8)W. K. Bratton. I. Szilard, and C. A. Cupas, J. Org. Chem., 32, 2019 (1967). (9) R. S. Butler, A. S. Carson, P. G. Laye, and W. V. Steele. J. Chem. Thermodyn., 3, 277 (1971). (IO) A. S. Carson, P. G. Laye, W. V. Steele, D. E. Johnston, and M. A. McKervey, J. Chem. Tbermodyn., 3, 915 (1971): T. Clark, D. E. Johnston. H. Mackle, M. A. McKervey, and J. J. Rooney, J. Chem. SOC., Chem. Commun., 1042 (1972).

Timothy Clark, Trevor McO. Knox, Henry Mackle* M. Anthony McKervey,* John J. Rooney Departnienr of Chemistry, The Queen’s University Belfast BT9 5AG, Northern Ireland Received March 18, I975

On the Isomerization of Pentacoordinated Molecules Sir:

The isomerizations of a pentacoordinated molecule, MABCDE, with five different ligands have been extensively discussed. Unless all six atoms a r e coplanar, such a molecule can have no symmetry element other than the identity. Several authors’ have pointed out that interconversions among the 20 possible isomers can be classified formally in terms of five different types or “modes” of rearrangement: two axial ligands exchange (aa), an axial and an equatorial ligand exchange (ae), an axial and two equatorial ligands exchange cyclically (aee), two independent axial-equatorial exchanges take place simultaneously (aesae), or two axial and two equatorial ligands exchange cyclically (aeae). Both the Berry mechanism2 and the turnstile mechanism3 lead to the mode aeae, but the view appears to be held’ that the other modes must be seriously considered. At the risk of laboring the obvious we wish to point out here that the first four modes a r e only to be considered as formal possibilities and may safely be excluded in any discussion of the detailed mechanisms. This is clear from a Journal of the American ChemicalSociety

/

97:13

/

Figure 1. (1-3): an aeae rearrangement proceeding in a single step through a pentahedral intermediate. (6-S), an aee rearrangement proceeding through a single intermediate. Since the intermediate has two quadrilateral faces sharing two edges it must be planar. (1-5), the same aee rearrangement proceeding through two aeae steps.

consideration of the topological properties of the coordination p ~ l y h e d r a ;that ~ is, we ignore metrical aspects. W e may choose the vertices of the coordination polyhedra either a t the atoms bonded to M or a t the ends of the five unit vectors that originate a t M and lie along the bond directions. Either way, only three topologically distinct kinds of polyhedra are possible: (a) trigonal bipyramid or hexahedron (six triangular faces); (b) tetragonal pyramid or pentahedron (four triangular faces and one quadrilateral face); (c) the degenerate polyhedron with all five vertices in a plane.4 Since there is no symmetry we shall refer to these as the hexahedral, pentahedral, and planar forms to avoid the symmetry connotations associated with the more familiar names. If the five vectors from M to the vertices A-E remain distinct, that is, no two bond directions ever coincide in the course of the rearrangement, then clearly any interconversion of one hexahedral isomer into another has to pass either through an intermediate pentahedral form or through an intermediate planar form. This follows because the edges that are not common to the polyhedra of the initial and final states must disappear, that is, become diagonals of planar faces, in the intermediate. Single-step mechanisms involving an intermediate pentahedral form are associated exclusively with the aeae mode (Figure I ) ; hence any single-step mechanism associated with the first four modes must involve an intermediate planar form. The first four modes can, of course, correspond to multistep processes involving pentahedral intermediates, but this is equivalent to replacing them by a sequence of aeae steps. An example is shown in Figure 1. The conclusion we draw from this is that while five rearrangement modes are formally possible, the chemical improbability of planar intermediates eliminates four of these a s serious possibilities, leaving only details of possible aeae mechanisms (Berry and turnstile, for example) to be explained further. Since a pentahedral form must occur somewhere along the reaction path for all chemically feasible rearrangements, it would appear that the most economical procedure for mapping the reaction pathway by quantum mechanical calculations might be to find the most stable pentahedral form and to proceed from there. Exactly analogous arguments apply to interconversions between the 30 possible pentahedral isomers, which must proceed either through hexahedral intermediates or through planar ones.

References and Notes (1) Some references are: (a) E. L. Muetterties, J Am. Chem. Soc., 91, 1636, 4115 (1969); (b) M. Gielen, J. Brocas, M. De Clercq, G. Mayence. and J. Topart, 3rd Symposium on Coordination Chemistry, Debrecen, Hungary, 1970, VoI. 1, pp 495-505: (c) J. Musher, J. Am. Chem. Soc., 94, 5662 (1972), J. Chem. Educ., 51, 94 (1974); (d) W. G. Klemperer. J. Am. Chem. SOC., 94, 6940 (1972); (e) W. Hasselbarth and E. Ruch, Theor.

June 25, I975

3837 Chim. Acta, 29, 259 (1973);(f) A. T. Balaban, D. Farcasiu, and R. Bani-

ca, Rev. Roum. Chim., 11, 1205 (1966):(9) A. T. Balaban, ibid., 18, 855

Table I. Partial 'H NMR (100 MHz) and Deuterium (D) LabelingQ Analysis of Id and 2d

(1973):Reference f actually discusses a related problem concerning the Wagner-Meerwein rearrangement.

-

H3

H1, '

(2)R. S.Berry, J. Chem. Pbys., 32, 933 (1960). (3) I. Ugi. D. Marquarding. H. Klusacek, P. Gillespie, and F. Ramirez, ACC. Chem. Res., 4, 288 (1971). (4)D. Britton and J. D. Dunitz, Acta Crystallogr., Sect. A, 29, 362 (1973).

r

Doyle Britton* Department of Chemistry, University of Minnesota Minneapolis, Minnesota 5.5455 J. D. Dunitz* Organic Chemistry Laboratory, Swiss Federal Institute of Technology, 8006 Zurich, Switzerland

A Remarkably Facile 1,3-Sigmatropic Suprafacial Shift with Retention of Stereochemistry. Catalysis of Carbon-Carbon Bond Migration by an Amide Ion Substituent Sir.

Since Woodward and Hoffmann' in 1965 first announced the principle of the conservation of orbital symmetry in concerted reactions, there have been refinements of the theory and the presentation of alternative schemes for discussion of pericyclic reaction^.^,^ One modification or clarification of the Woodward-Hoffmann rules1 involves the allowed stereochemistry of 1.3 shifts, where it appears two concerted pathways can be a ~ a i l a b l e .A~ ~fine . ~ balance of opposing electronic and steric effects suggests that when migration by an electronically favored suprafacial inversion (si) pathway is sterically prohibited, the suprafacial retention (sr) pathway can proceed in a concerted manner. Herein we report a novel example of the sr 1,3 shift5s6 in which a dramatic effect on reaction energetics results from hydrogen atom removal on a nitrogen substituent adjacent to the migrating enter.^ The N-benzyl7 ( I b ) and N-carbomethoxy8 (Id) derivatives of 3-azabicyclo[3.3.2]deca-6,9-diene( l a ) have been purified by G L P C . Therefore, it was surprising when introduction of amine l a into a G L P C injector port heated to 250' (column temperaturelo 120') was followed by isolation of only amine 2a" (90% recovery), the product of 1,3sigmatropic rearrangement. Even more surprising, if amine l a in either diethyl ether or n-pentane was treated with methyllithium a t 30' after 1 min a complete conversion to amine 212 (90% recovery) resulted.

la. R = H

b, R = CH?Ph c, K=COOPh d. R = C O O M e

2a. R - H

b, R

=

CHLPh

c . R = COOPh d. R = COOhle

e. R = COPh To study rearrangement stereochemistry amine l a was synthesized13 with one deuterium atom selectively distributed among the positions CY to nitrogen. Selectively labeled la was then rearranged to labeled 2a and the fate of the label during rearrangement was determined by ' H N M R integration (Table I). Because of conformational mobility of 1, dihedral angle relationships do not reveal whether H 2 / H 3 or H 4 / H 5 is the exo pair. However, the deuterated methylurethanes8 Id and 2d could be used to assign H 3 of Id, and

2d

Id

Id

Reaction path,f calcd %-one D for 2d

2d

Shift

Shifte

Proton

(6)b

D

H2 H3 H4 H5

2.7% 2.798 3.02ki 3.36hJ

35 i- 3 58 3.5 3.5

ra c sr (sr + si)

(6)

D (70)

4.09k (4.09)

92m(90)n

93

77

62

2.851

8m (10)n

7

23

38

si

(2.85)

QStandard deviations were used to determine the most probable error 6y/y = [(&z/u)' + (6b/b)' + ...I v2. 'H NMR of the phenyl carbamate derivatives I C and 2c confirmed these results. b Benzened,, 76". CH2 + H 3 = 9 3 i 370, the ratio (H2 + H3)/(H4 + H5) was determined via relative 'H NMR peak areas. d T h e H2/H3 ratio was determined from relative 'H NMR peak heights; J. A. Dale and H. S. Mosher, J. Am. Chem. Soc., 90, 3732 (1968). eCDCl,, 76". fsr, si = suprafacial retention and inversion, respectively, rac = racemization. g d , J = 1 4 Hz. hBroad. i h a d i a t i o n of H6 ( 6 2.51) collapses H4 t o a d , J = 1 4 Hz. iIrradiation of H1 (6 2.00) collapses H5 t o a d, J = 1 4 Hz. k d , J = 1 3 Hz. l d d , J = 4 Hz, 1 3 Hz. m H 2 + H3 = 9 2 i l?, from l a and methyllithium in ether or pentane. n H 2 + H 3 = 9 0 i 270, GLPC of l a at 250" injection temperature.

thence from coupling data all protons CY to nitrogen. First, the endo proton H 3 in 2d was assigned by its W-planI5 coupling with the bridge proton H7. Second, the exo proton H3 of Id has not undergone stereomutation during rearrangement so it has been transformed into endo proton H 3 of 2d. It then follows since 2d has 92% of the label a t the endo H 2 / H 3 positions that the most deuterated shift position of Id a t 6 2.79 (58% D) must correspond to exo proton H 3 . Comparison of the situs of deuterium label of Id determined by IH N M R integration with the predicted deuterium label for various stereochemical possibilities (Table I ) indicates net retention of stereochemistry (sr) during thermal or methyl lithium catalyzed sigmatropic rearrangement. The yield and stereochemical data, while consistent with a concerted mechanism, does not exclude stepwise ionic or radical pathways.I6 Error limits of the data do allow for contribution from random processes of 46% (77% sr, 23% si) during thermal rearrangement and 29% (85% sr, 15% si) during base catalyzed rearrangement of la. The most favorable overlap of the lone pair on nitrogen with the migrating u bond results when R is exo in the N anti conformation 1A and R is endo in the N-syn conformation 1B.I8 In both these conformations a substituent group, R = CHzPh or COOMe, is in the sterically more crowded position and should as observed raise the transition state energy relative to R = H . Removal of the nitrogen substituent

A

1A

Communications t o the Editor